What’s New

Project Update

For this newsletter, we wanted to “open the hood” on the many interconnected research components that ultimately come together to inform the City Food Policy Project’s (CFPP) agent-based model (ABM). While the ABM is the project’s central integrative analytical framework, it is built from a wide range of complementary analyses that help ensure the model reflects real-world conditions, behaviors, and tradeoffs. These inputs include qualitative interviews and focus groups with producers and supply chain actors, causal loop diagrams that translate stakeholder insights into system structure, game-theoretic economic bidding research that helps us understand how vendors may respond to different procurement rules, life cycle assessments to estimate environmental impacts of policy shifts, and extensive data assembly, cleaning, and validation across commodities and supply chain stages. We are also documenting this work through a public-facing QuartoBook and GitHub repository to support transparency, reproducibility, and future transferability to other cities. The updates below highlight several of these pieces to illustrate how the project’s interdisciplinary research approach is helping us build a more realistic and decision-relevant model of how values-based public food procurement policies may shape economic, environmental, and social outcomes in New York City and beyond.

Analysis of the Interviews and Focus Groups

This winter and spring, the CFPP team has been advancing the qualitative research components of the project, with a focus on understanding how New York State producers and mid-supply chain actors make decisions about whether, and under what conditions, to participate in urban public food procurement markets. Drawing on interviews and focus groups with growers, processors, aggregators, and wholesalers across the dried bean, beef (cull dairy), and leafy greens supply chains, the team synthesized key findings on the economic, relational, and operational factors that shape market participation. Our analysis indicates that participation decisions are influenced not only by price, but also by trusted business relationships, buyer reliability, logistical feasibility, infrastructure constraints, and deeply held personal and place-based values. These insights are helping to explain why even well-designed procurement policies may encounter implementation barriers, while also identifying opportunities to better align public purchasing strategies with supply chain realities.

Importantly, this work is directly informing the project’s modeling efforts. Findings from the interviews and focus groups are being used to populate and refine the decision-making functions within the project’s ABM, strengthening its ability to simulate how supply chain actors may respond to alternative values-based procurement policy scenarios in New York City. By incorporating behavioral insights related to capability, opportunity, and motivation, the model is becoming better equipped to reflect real-world decision-making rather than relying on simplified assumptions about price responsiveness alone. This component of the work is being led by Dallas Selle, an MPS student in Global Development at Cornell University, whose leadership has been instrumental in advancing the qualitative analysis and integrating it into the project’s broader analytical framework. Look out for a peer-reviewed journal article that summarizes these findings!

Causal Loop Diagrams

The team focused on synthesizing qualitative insights from interviews, focus groups, and participatory workshops into a set of causal loop diagrams (CLDs) representing each of the three commodity supply chains. These CLDs map the key feedback structures, delays, and interdependencies that govern system behavior, incorporating perspectives from producers, intermediaries, and public procurement stakeholders. By formalizing how actors perceive relationships among prices, demand, capacity constraints, and policy incentives, the diagrams make explicit the mechanisms through which proposed NYC procurement policies, such as local sourcing requirements, organic premiums, and infrastructure investments, may influence supply chain dynamics. This process has also helped clarify system boundaries for the ABM and identify leverage points where interventions may be most likely to produce meaningful change.

The CLDs now serve as the conceptual foundation for the each commodity model. They are being used to guide model structure, decision logic, and the representation of interactions among producers, processors, wholesalers, and public buyers. In this way, the CLDs provide a systematic bridge between stakeholder-informed qualitative knowledge and formal simulation, helping ensure that the model is grounded in observed dynamics while still enabling rigorous analysis of policy tradeoffs.

Bidding

The team, led by Frank Ge from the Dyson School of Applied Economics and Management at Cornell University, advanced a stream of research focused on how vendors may respond to values-based public food procurement bids when cities incorporate attributes such as local sourcing or M/WBE participation into food solicitations. Using a novel game-theoretic (Nash equilibrium) bidding framework, the team is analyzing how vendors adjust bid prices under different procurement environments, with particular attention to how product attributes, vendor competition, and policy design influence bidding behavior. The current paper uses lettuce as a focal example and is calibrated using New York City and New York State procurement data, allowing the team to examine how specific procurement rules, such as NYC’s ability to consider qualifying bids priced up to 10% above the low bid, may shape both bid prices and vendor participation. Early findings suggest that integrating values into bid requests can meaningfully alter prices: for example, local products may carry price premiums that approach the 10% threshold at which NYC can legally consider them, while M/WBE set asides can increase bid prices if they reduce the size of the competitive pool. The work also highlights the importance of vendor participation, showing that additional bidders can substantially reduce prices, although excessive competition may also reduce supplier viability over time.

This bidding research is serving as a key input into the broader City Food Policy agent-based model. While the bidding paper is a stand-alone methodological contribution, its results are also being used to inform and validate how the model represents vendor pricing behavior and procurement responses under alternative policy scenarios. This is especially important because the success of values-based procurement depends not only on supply chain capacity, but also on how vendors strategically respond to changing bid structures and incentives. By grounding these behavioral assumptions in both theory and empirical NYC procurement data, the team is strengthening the model’s ability to simulate more realistic market responses to policy changes. In turn, this improves the project’s broader capacity to estimate whether different procurement strategies are likely to be feasible and achieve New York City’s values-based goals.

Life Cycle Assessment (LCA)

The team completed a preliminary life cycle assessment (LCA) of the project’s policy actions to estimate how proposed procurement changes may influence the environmental footprint of beans procured by New York City under different scenarios. The LCA system boundary (i.e., what is included in the analysis) includes production, processing, and distribution. The analysis incorporates changes in farmer-level input use (e.g., fertilizer, manure, pesticides, and seed) for both local and non-local production systems and translates those differences into environmental impact estimates using established methods. Indicators include greenhouse gas emissions, water quality (nutrient losses), resource use (land, water, and energy), and input-specific measures such as fertilizer, manure, and pesticide use.

One focal area of this work is the project’s “menu change” policy action, which is operationalized as a shift in the composition of protein procurement by NYC agencies. We’re examining the relative shares of bean proteins compared to beef-based proteins, and if they are locally (New York State) or nonlocally grown or raised. Empirical observations of recent procurement trends are being used to parameterize a business-as-usual scenario (i.e., what New York City is currently purchasing), while counterfactual scenarios represent expanded adoption of bean-based proteins and increased local sourcing. These modeled shifts are then linked to life cycle inventory data to estimate marginal changes in environmental impacts attributable not only to dietary substitution, but also to changes in the geographic configuration of supply chains.

Agent-Based Model (ABM)

The team completed the initial build of the dry bean ABM and began validation (i.e., confirming with stakeholders that the model outputs are reasonable). Using the CLDs as the conceptual foundation, the model represents all major supply chain participants relevant to the dry bean system, including producers, aggregators/cleaners, primary and secondary processors, wholesalers, NYC agencies, and external commodity and differentiated markets. The model boundaries capture the production, processing, distribution, and procurement pathways most likely to respond endogenously to policy change.

The team assembled data inputs from restricted-access Census of Agriculture, crop budgets, interview and focus groups, NYC purchasing records, commodity and supply chain price information from the US Department of Agriculture, LCA inventory information, etc. These data were used to parameterize the model with key inputs such as production costs, margins, supply chain attributes, and agent behavior. Agents interact through contracting, purchasing, and market transactions that determine prices, commitments, and product flows. Importantly, the model’s behavioral assumptions are grounded in the qualitative insights generated through interviews, focus groups, and participatory workshops / validation sessions, allowing the team to translate stakeholder-informed causal structure into formal rules governing production, contracting, purchasing, and responses to procurement incentives. This creates a robust bridge between qualitative evidence and simulation-based evaluation of values-based public food procurement policies.

The beef ABM is now well under way and will be completed in June 2026, with the greens ABM to follow.

Validation

The team also began the model validation process, which is designed to assess both technical performance and policy relevance. Initial work has focused on verifying that model logic, equations, parameter values, and agent decision rules are internally consistent and correctly implemented relative to the conceptual model, causal loop diagrams, and source data. The team has also begun testing behavioral validity by assessing whether the model reproduces expected system patterns under baseline conditions and responds plausibly to policy shocks, including sensitivity analyses to identify which assumptions and parameters most strongly influence results.

Next steps in validation will include stakeholder review and participatory validation with producers, intermediaries, and NYC policy partners, allowing the team to compare simulated behaviors and outcomes against lived experience and expert knowledge. The final stage will focus on policy relevance, evaluating whether the model generates transparent, interpretable, and decision-useful estimates of tradeoffs across environmental, economic, and procurement outcomes under alternative scenarios.

Dashboard Updates

The team made substantial progress in developing and refining the project’s public-facing interactive dashboard, a key translational output designed to make the project’s modeling results more accessible and useful to four primary user groups: procurement staff responsible for bids, contracts, and menu planning; food policy analysts such as academic researchers and policy-focused institutions; food policy decision-makers including government actors such as the NYC Mayor’s Office of Food Policy and the New York State Department of Agriculture and Markets; and practitioners such as organizations working to influence food, agriculture, and nutrition policy. To support this effort, the team hired a specialized developer to build a web-based dashboard that allows users to explore selected economic and environmental tradeoffs associated with New York City’s values-based public food procurement strategies. Importantly, the dashboard is not the model itself. The underlying ABM generates hundreds of outputs. The dashboard instead highlights a curated subset of indicators most relevant to New York City and New York State policy goals and values, allowing users to engage with the project’s findings through a more intuitive, decision-relevant interface. The dashboard supports users to examine six food policy actions across our three focal commodities, beans, beef (cull dairy), and salad bar greens (leafy greens and cabbage), and to compare how different policy definitions and implementation approaches may affect multiple stakeholder groups.

To learn more about the dashboard, please visit: https://www.cityfoodpolicy.com/dashboard/

Looking Beyond New York

Transferability

Over the past year, the CFPP team has advanced a major body of work focused on understanding the international transferability of values-based public food procurement strategies across diverse urban contexts. This effort began with a series of semi-structured interviews with researchers, practitioners, funders, and city representatives from North America, Latin America, Europe, and Africa to better understand how cities are defining, implementing, and evaluating values-based procurement. Building on those interviews, the team then convened two international workshops designed to deepen cross-city learning. In August 2025, the team partnered with the City of Copenhagen to host a workshop on Accelerating Values-Based and Regenerative Food Procurement, bringing together municipal staff, researchers, NGOs, city networks, and funders to examine the experiences of New York, Nairobi, Barcarena, Milan, and Copenhagen. In October 2025, the team extended this work through a parallel session at the Milan Urban Food Policy Pact Global Forum, where participants worked through nine additional city cases from across Europe, North and South America, Africa, and Asia. Across both convenings, participants identified common policy approaches, mapped who is directly and indirectly affected by procurement interventions, and surfaced the information gaps that most constrain effective design and implementation.

Together, these interviews and workshops helped the team better understand which elements of values-based procurement are broadly transferable across cities, and which must be adapted to local supply chain, governance, and institutional conditions. Early findings suggest that while cities differ substantially in their goals and contexts, there are important commonalities in the policy tools they are using, such as dietary guidelines and menu changes, purchasing requirements, contract preferences, vendor incentives, and infrastructure investments, as well as in the enabling conditions needed for success, including stakeholder engagement, intra-governmental coordination, and stronger supply chain mapping. This comparative work is helping the team identify where the CFPP’s methodological approach can be adapted to other cities seeking to use public procurement as a lever for economic, social, and environmental goals. A journal article by Jill Clark, Thomas Forster, Annie Trevenen-Jones, and Becca Jablonski is being developed as one output of this work and will synthesize these findings to help document how values-based procurement is being implemented across contexts and where major evidence gaps remain.

Where to Find Us

In the coming months the CFPP Team will be sharing its work and participating in conversations that address values-based public food procurement, supply chain development, and urban-rural interdependence.

May 5-6, Chicago, IL, Google Food Lab, CFPP Supply Chain and Communications Lead Michael Hurwitz (Landing Lights Strategies, Inc.) will join a collaborative community of industry leaders, change agents, and visionaries to discuss and share strategies to disrupt the status quo and create positive and transformative food systems change.

May 27-28, Rome, Italy, Forum Compraverde Buygreen, https://www.forumcompraverde.it/en/program-2026-may-the-27th/

May 25-28, Rome, Italy, Rome Nutrition Week 2026, https://www.unnutrition.org/rome-nutrition-week-2026

June 7-10, Red Bank, NJ, Northeastern Agricultural and Resource Economics Association Annual Meeting, https://narea.org/meeting-2026/

July 26-28, Kansas City, MO, Applied Agricultural Economics Association Annual Meeting, https://www.aaea.org/meetings/2026-aaea-annual-meeting

July 19-23, Porto Alegre, Brazil. World Congress of Rural Sociology. CFPP advisors Juliana Tangari and Annie Trevenen-Jones and CFPP team member Thomas Forster will present on ““Emerging Practices of Values-Based Public Food Procurement for Integrated Urban and Rural Development””.

Additionally, CFPP advisor Letitia Campos Baird will present on the Escola Sustentável (Sustainable School) initiative. This selection highlights the success of her and her colleagues’ efforts in strengthening local smallholders and promoting sustainable public procurement and serves as a significant international validation of the impact their institutional strategies are having on regional development.

In our next What’s New, CFPP will share the Dashboard, opportunities for feedback, and key lessons learned. Please contact us if you are interested in a personalized Dashboard demonstration and make sure to check out our team members’ latest journal articles in the What We’re Reading section of the site. Follow this link to receive regular updates about CFPP.